Accurate microscopic identification of human spermatozoa is important in sexual assault cases. We have compared the results of examinations with (1) a fluorescent microscopy method, SPERM HY-LITER™, and (2) Baecchi's method for identification of human spermatozoa. In 35 artificial, forensic type samples, spermatozoa were identified in 45.7% with SPERM HY-LITER™ in Copenhagen, in 54.3% in the laboratory of the manufacturer of SPERM HY-LITER™, and 40.0% of the samples with Baecchi's staining method. When differences occurred between the two methods, it was significantly more often that SPERM HY-LITER™ detected spermatozoa when Baecchi's method did not (ts=6.567, df=1, P=0.048). This trend was also seen in selected compromised or degraded samples and in selected adjudicative samples. The reactions with spermatozoa from dog, horse, pig and bull were negative with SPERM HY-LITER™, whereas Baecchi's method was non-selective. Data from forensic casework samples in Copenhagen from two years (2008 and 2009) are presented. The samples from 2008 were investigated using Baecchi's method, while those from 2009 were investigated using SPERM HY-LITER™. The frequencies of positive results were similar between the two methods for the two years (27.9% and 32.1% respectively). Analysis of acid phosphatase (ACP) activity for the positive results obtained for these two years does not support the use of a negative ACP result as a prescreen for microscopic analysis for spermatozoa.