Herpes simplex virus type 1 (HSV-1) is a ubiquitous human pathogen. The a sequence of HSV-1 is the cis-acting site required for the cleavage and encapsidation of unit-length HSV-1 DNA from concatemeric forms. The consensus a sequence consists of (i) DR1 (direct repeat 1), (ii) Ub, (iii) a DR2 array [a repeat of various copy numbers of DR2 elements (11 or 12 bp)], (iv) a DR4 stretch and (v) Uc. In the present study, the nucleotide sequences of the a sequences of 26 HSV-1 isolates were determined and the DR4 stretches were classified into three groups. The state of a set of 20 DNA polymorphisms in the genomes of these HSV-1 isolates was determined previously. A correct classification rate of 100 % was achieved when discriminant analysis was performed between the DR4 stretch (criterion variable) and the set of 20 DNA polymorphisms (predictor variables), suggesting a close association of the DR4 stretch with HSV-1 diversification. DR2 elements of 9, 13 and 14 bp were detected in addition to those of 11 and 12 bp, and a correct classification rate of 93 % was achieved when discriminant analysis was performed between the DR2 array and the set of 20 DNA polymorphisms. Some DR2 elements of one HSV-1 isolate had the same nucleotide sequences as part of the adjacent DR4 stretch, and these variations were adequately explained by postulating recombination involving DR2 elements; hence, the DR2 array was deduced to be prone to recombination.