The relationship between the three basic parameters of ejaculated spermatozoa, i.e. concentration, motility and morphology, and the results of intracytoplasmic sperm injection (ICSI) were investigated in 838 microinjection cycles. A further 123 ICSI treatment cycles in which testicular spermatozoa were used for microinjection were also evaluated. The influence of anti-sperm antibodies (ASA) on the outcome of ICSI was investigated by analysing 55 cycles where the proportion of ASA-bound spermatozoa was > or =80%. After microinjection, oocyte intactness, fertilization, embryo cleavage, transfer and pregnancy rates were recorded and compared. The results showed that neither the type nor the extent of sperm impairment had an important influence on the outcome of ICSI when ejaculated spermatozoa were used. Only two very rare conditions had a strongly negative influence on the result of ICSI, i.e. where immotile (presumably dead) spermatozoa or where round-headed spermatozoa were injected into the oocyte. Neither the proportion of ASA-bound spermatozoa, the type of dominantly present ASA, nor the location of ASA on the spermatozoa had an important influence on fertilization, embryo development or pregnancy rates after ICSI. In most of the cycles combined with testicular biopsy (79%), there were enough motile spermatozoa present in the wet preparation for injection of all the oocytes. Injection of motile testicular spermatozoa led to a higher normal fertilization rate than did injection of non-motile spermatozoa (65 versus 21%). It can be concluded that injection of motile (living) spermatozoa into oocytes is the most important factor in determining good results with ICSI and that other sperm parameters do not have a strong influence on the outcome of ICSI.