Efficacy of drug detection by fully-trained police dogs varies by breed, training level, type of drug and search environment

To best understand this article in the context of the Breeds and Behavior literature, please see National Canine Research Council’s complete analysis here.

Jezierski, T., Adamkiewicz, E., Walczak, M., Sobczyńska, M., Górecka-Bruzda, A., Ensminger, J., & Papet, E. (2014). Efficacy of drug detection by fully-trained police dogs varies by breed, training level, type of drug and search environment. Forensic Science International237, 112-118. doi: 10.1016/j.forsciint.2014.01.013

National Canine Research Council Summary and Analysis:

Breed differences were not the primary focus of this study, but the paper is included in the discussion because differences between breeds were found, and the methodology with regard to breed differences demonstrates how potentially confounding variables make these conclusions suspect. (See also Hall et al., 2015 for a fuller consideration of these variables in attempting to determine any breed-based aptitude differences in scent work in general).

Subjects were 3 specific breeds: 68 Labrador Retrievers, 61 German Shepherds, 10 English Cocker Spaniels, and 25 dogs from various breeds categorized as members of the terrier group. The “Terrier” group makes the results difficult to evaluate as it was comprised of various breeds (at least 4 although this is not altogether clear) of varying historical functions and no particular genetic relatedness. All the dogs were considered by Polish police training protocols to be fully trained drug detection dogs, but some were tested for this study prior to their first certification. The authors do not indicate how breed was determined, though they do note that the Labrador Retrievers were purebred. A total of 1,219 drug odor detection search trials were conducted (440 with German Shepherds, 517 with Labrador Retrievers, 203 with “Terriers”, and 59 with English Cocker Spaniels). The handler was blind to the target’s location, but the experimenter was not. The experimenter filmed sessions for later analysis. Dependent variables included the number of correct trials, latency to find the target, number of times the target was passed within 1m without a signal, number of false signals, and number of timed-out trials. Targets were a variety of drugs (marijuana, hashish, amphetamines, cocaine, and heroin), and tests were conducted in familiar and unfamiliar rooms, outdoors, through a lineup of luggage, and inside and outside of cars.

The German Shepherds had a slight, but significantly higher total percentage of correct indications than the Labrador Retrievers; the difference was not significant between the German Shepherds and the English Cocker Spaniels. The biggest difference was found between the “Terriers,” which had significantly lower percent of correct indications when compared to each of the three breeds.  However, since this comparison involved lumping together at least 4 unrelated breeds, the likely sample size of each of these breeds is probably too small to give credence to this finding. Potential confounds (beyond the questionable breed grouping) are the impossibility of knowing how much the trainer’s expertise affected the dog’s competence and how many of the dogs never actually passed their certification.

Abstract and Link to Purchase Full Text of the Original Article:

Additional Reference:

Hall, N. J., Glenn, K., Smith, D. W., & Wynne, C. D. L., (2015). Performance of Pugs, German Shepherds, and Greyhounds (Canis lupus familiaris) on an odor-discrimination task. Journal of Comparative Psychology, 129(3), 237-246. doi:

Page last updated July 16, 2019