Background Questionable systolic assistance and latissimus dorsi (LD) muscular degeneration as a result of continuous electrical stimulation constitute important drawbacks to dynamic cardiomyoplasty. To avoid full transformation of the LD and thereby cause better systolic assistance, a new stimulation protocol was developed. Fewer impulses per day are delivered so that the LD wrap has daily periods of rest (demand), based on a heart rate cutoff. We describe our experience of demand dynamic wrapping by discriminating between patients with active systolic assistance and those with a passive girdle effect (adynamic-girdling). Methods Fourteen patients with primary dilated cardiomyopathy (13 men, 1 woman; mean age, 58.2 ± 5.8 years; 12 sinus rhythm, 2 atrial fibrillation) underwent dynamic cardiomyoplasty between 1993 and 1996 as well as the demand protocol at different intervals. Clinical, echocardiographic, mechanographic, and cardiac invasive assessment records, as well as cardiovascular events (death and arrhythmias), were retrospectively reviewed. The patients were divided into two groups on the basis of the mechanographic measurement of speed of contraction of the heart wrap, as measured by tetanic fusion frequency analysis before starting demand stimulation: demand dynamic wrapping patients with fast LD (high tetanic fusion frequency, 7 patients), and adynamic-girdling patients with slow LD contraction times (low tetanic fusion frequency, 7 patients). It was assumed that in adynamic-girdling patients dynamic assistance was virtually absent, so the wrapping acted only as a passive constraint wall. Results The two groups were comparable for sex, age, dilated cardiomyopathy cause, New York Heart Association class, and left ventricular ejection fraction at the start of the demand protocol period. After a mean duration of follow-up of 41.4 ± 21.1 months (range, 23 to 69 months), the demand dynamic wrapping group showed improved New York Heart Association class (1.14 ± 0.34 versus 2.07 + 0.18; p = 0.0004), higher values of left ventricular ejection fraction (34.6 ± 8.0 versus 26.5 ± 3.1; p = 0.005) and LD wrap tetanic fusion frequency (38.3 ± 5.88 versus 24.3 ± 2.93; p = 0.002), and a better survival (85.7% versus 28.6%; p = 0.037) than the adynamic-girdling group. Conclusions Demand dynamic wrapping offers good results in terms of fewer cardiovascular events and greater survival. When compared with the passive constraint effect of LD muscle, demand dynamic wrapping proved to be more effective.