Death records are a staple of genealogical research. They serve as primary sources for information about when and where an ancestor died and was buried. They also offer clues about the birth of the ancestor, his or her parents, the cause of death, and sometimes the spouse and occupation. Because most statewide indexes are broken down by years, these genealogical gems can stay buried, however, when an exact date of death is unknown. The problem is compounded when working in a time period when the records are held only at the county level, thus requiring the knowledge of where the ancestor died.

In the past, examining the indexes when the date or place of death was unknown has required researchers to do their research on-site, obtain the microfilm/fiche index through a local Family History Centers, or pay a research fee to the agency that held the index or to a professional researcher. While all of these methods work, the advent of the Internet has made the search a bit easier in several states. Let’s look at examples of online, statewide indexes from three states: Michigan, Ohio, and Kentucky.

The search engine is simple to use. The researcher can enter as much or as little information as desired. There is a limit of 200 results from a search, so searches for common surnames should be narrowed down by entering additional search criteria, such as county of death, or by limiting the search to just a few years at a time.

Search parameters can contain information entered into any or all of the following fields: decedent’s last name, decedent’s first name, father’s last name, county of death, and year of death. To help researchers avoid entering incorrect spellings for the counties or entering years outside of the index, those fields are controlled by drop-down menus. These menus contain an arrow on the right side of the field; clicking on the arrow reveals a list of acceptable values.

The use of "wildcards"–special characters entered to act as a variable in a search–is not supported. However, the search engine will give results for all surnames beginning with the letters that are entered. For example, a search on CHARB as the decedent’s last name yielded Charbenau, Charbeneau, Charbinau, Charboneau, Charboney, Charbott, Charbraneau, and Charbvuena.

Using the father’s surname as a search criterion can be both a help and a hindrance. It can be helpful in tracking down married daughters. However, just like all death records, the accuracy depends on the knowledge of the informant. There will be no matches if the father’s surname is in the database as "unknown" or "not recorded."

When searching for a surname that begins with a prefix, such as Mc, Mac, or De, it may be necessary to do two searches. Entering Mc Coy (with a space between the two Cs) yields different results than searching under McCoy (with no space).

The search results screen gives the deceased’s name, date of death, father’s surname, and county of death. Although this is wonderful information, it is only the beginning of the information. Each entry in the index is hyperlinked to an abstract of the death record.

The abstracted death record includes the ledger and page number where the entry was found. This can help the researcher obtain a photocopy of the record, either from the county or by ordering the microfilm from the Family History Library.

Other fields in the death record abstract include the deceased’s name, place and county of death, sex, race, marital status, age, cause of death, birth place, occupation, father’s name and residence, and mother’s name and residence. The date of the record is also included, which can aid researchers in evaluating the information contained within. The abstract notes when information was not given in a particular record.

The online index was created by scanning the original indexes and using OCR (optical character recognition) software to convert the digital images to text. The problem with this process is that some characters are not recognized properly. If the original index was not printed clearly, an O can become a C, a D can become an O, an E can become an F or a B, along with dozens of other possibilities.

According to the OHS Web site, the accuracy rate is estimated to be approximately 85 to 95 percent. Efforts are underway to compare the database with the original, printed index. Researchers are also encouraged to report errors they see in the database.

There are two search engines for this index. The basic search, which is the one located at the URL noted above, is the best place to start. Search fields include the deceased’s first name, last name, and county of death and the years to be searched. The default setting for county is "All Counties." Unless the surname being searched for is extremely common, leaving the county setting at "All Counties" can be effective, as it includes counties that are misspelled due to OCR errors.

The advanced search engine, located at ohiohistory.org, or found by clicking on the "Advanced Search" link from the basic search screen, contains many more fields to search on. In addition to those on the basic search, researchers can enter values for month of death, date of death, volume number, and certificate number.

The database truncates (cuts off) the first name after the seventh letter; last names are truncated after the 11th letter. Fortunately, neither search engine will allow the researcher to enter more characters than that.

Wildcard characters * and % can be used. Searches using H*NSON in Franklin County from 1913 to 1917 yielded Hackinson, Hoskinson, and Hutchinson. The search engines will also find any entries beginning with the letters entered. Searching for JOHNS in Franklin County from 1913 to 1917 yielded Johns, Johnson, and Johnston.

The results screen gives the deceased’s name, county of death, date of death, and volume and certificate number. After the correct entry is found, a copy can be ordered from the Ohio Historical Society. It is important to cite the exact record requested, as Ohio law requires copies to be made at cost, but allows for a research charge when the exact record is not known. In other words, it is cheaper to look it up yourself and then send for a copy than it is to send in a request for a search and the copy.

The current copy fee is $1 per request, which includes up to four copies, with additional pages being 25 cents each, plus excess postage. The base fee for research starts at $5 for one name. Details are available at the Web site. As with any repository, fees are subject to change.

The death records search engine for 1911—86 will allow the researcher to enter a value for the first and/or last name as well as the place of death or place of residence. It is not possible to search using both the place of death and the place of residence as search criteria. There is a relatively low limit on the number of results allowed from a search. When searching for KIRK, the search stopped at 35 hits. A search for JOHNS stopped at 36. Therefore, narrowing the search by location may be necessary. However, the search engine for 1987—92 allows searches by name only. With this database containing a smaller number of records, it may not be as difficult to find the correct record.

Using the place of death may be tricky, as the database uses codes to represent the place of death. A list of these places can be found by following the links from the search screen. Narrowing the search by place of residence is even more difficult due to the number of variant spellings in the database. For example, there are five different spellings for Brandenburg.

The search engine will not automatically give all surnames beginning with the letters entered. However, searches using ? or * as wildcards are allowed. Using JOHNS* as the name, with the place of death as Floyd, the results included people with surnames of Johns, Johnson, and Johnston before the search results reached its limit. A second search, still using Floyd as the place of death, but using JOHNS? as the name gave results for people with the last name of Johns, plus many men with the first name of John. When conducting research for this article, I was not successful in using wildcard characters in the place of death or residence fields.

The results of the search include the name of the deceased, date of death (in YY-MM-DD format), age, place of death, place of residence, volume number, and certificate number.

An added feature of this database is the hyperlinks from some places listed in the results screen. For example, from the result screen showing Albert Kirk, who died in Pike County, the researcher can click on Pike and be taken to the Pike County page of the Kentucky Atlas and Gazetteer Web site, hosted by the University of Kentucky. This page contains a map of the state showing the county’s location, gives a brief overview of the county, and includes a search engine to find additional features in the county.

Sometimes less is more. Entering too much information can eliminate records that may be of use to the researcher. For example, searching on Katharine Fannan will eliminate any records for Katherine Fannan, Katie Fannan, or Mrs. John Fannan. A search for George Skinner will eliminate any records for Geo. Skinner or G. W. Skinner. Keeping an open mind when using these databases will help yield the necessary records.

The Internet is an excellent medium for sorting through large amounts of data without having to go on-site or rent microfilm. These three online, statewide death indexes are a prime example.

Amy Johnson Crow, CG, is a professional genealogist who specializes in Ohio research. She is a lecturer, past editor of the Ohio Civil War Genealogy Journal, recording secretary of the Ohio Genealogical Society, and chairperson of First Families of Ohio. A member of NGS, APG, and numerous state and county societies, she can be reached at amy@amyjohnsoncrow.com.