The modified zinc test is evaluated as a screening instrument in forensic semen identification using the classical acid phosphatase test as a "gold standard" reference. Both tests were applied to vaginal swabs taken from women (n = 456) at varying time periods after sexual intercourse and the corresponding sensitivity and specificity were calculated for both tests separately and for their parallel and serial combination strategies, with the microscopical visualization of sperm as end criterion. The results were submitted to Bayesian calculations and a model was obtained, giving the posterior probabilities of finding or not finding sperm in a microscopical preparate for any possible prior probability. For the largest part of the theoretical sperm prevalence range, negative predictions were more informative than positive ones. With the prior probability of finding spermatozoa set at P = 0.5 (Bayes' postulate of neutral prior probability), the model was applied to a set of 192 pieces of evidentiary material from alleged sexual assaults.