To the Editor: There has been pronounced variation in the incidence of neuroleptic malignant syndrome over the last 20 years. Neuroleptic malignant syndrome was considered rare in the 1960s and 1970s until several retrospective studies were published. Thereafter, interest intensified, and reports of "atypical" formes frustes and incipient neuroleptic malignant syndrome (none of which met diagnostic criteria) appeared in the literature R5215511CHDCDAED. However, later prospective studies, which cited incidence rates for neuroleptic malignant syndrome between 0.07% and 0.15%, confirmed that this is indeed a rare (but certainly not-to-be-overlooked) condition. The recent decline in incidence may, in part, be attributable to more judicious use of conventional antipsychotic medication (e.g., cessation of rapid neuroleptilization, better hydration of patients during titration of medication) R5215511CHDDHEJC. In addition, heightened clinical awareness, careful investigation to rule out alternative diagnoses, and greater diagnostic specificity are also major contributory factors. The recent report by Newman and colleagues R5215511CHDDFGGC of atypical neuroleptic malignant syndrome with risperidone adds to an accruing literature on the syndrome with the use of novel antipsychotics but, in our opinion, complicates further this diagnostic conundrum. The spectrum concept of neuroleptic malignant syndrome, dubious at best, is particularly tenuous in treatment with novel antipsychotics R5215511CHDEBHDA. Several of the side effects typically observed in the initial titration period with novel antipsychotics (autonomic dysregulation; benign hyperthermia with clozapine) resemble manifestations of neuroleptic malignant syndrome R5215511CHDEBHDA. In another example, a recent prospective study of 37 patients receiving clozapine reported elevations of creatine phosphokinase (range: 725–20,000 IU/liter) in 29 patients who were without any other evidence of neuroleptic malignant syndrome R5215511CHDBJBHJ. Collectively, these observations should be cause for caution in hastily ascribing a diagnosis of neuroleptic malignant syndrome during treatment with novel antipsychotics. At the present time, our understanding and diagnostic specificity for neuroleptic ma­ignant syndrome is too rudimentary to advance the notion of a clinical spectrum, particularly in the context of atypical antipsychotics.