A unique mechanism of "cross-talk" has recently been described in the Gemini 415A DDD pacemaker in which cross-inhibition of the ventricular output stimulus may occur in the early Gemini 415A models which were manufactured with dual anodal rings interconnected by a wire. The Gemini 415A units with dual anodal rings were implanted under investigational protocol in 53 patients from September 1982 to August 1984. To date, cross-talk has been discovered in six patients during clinical follow-up, manifesting as inappropriate ventricular inhibition. Syncope was documented in one patient. Non-Cordis atrial and ventricular leads were present in all six patients and five patients had at least one chronic lead. Cross-talk was neither evident nor provokable in Gemini 415A units which had compatible Cordis lead systems. Reduction of the atrial output and/or the ventricular sensitivity prevented cross-inhibition in three patients. Cross-talk persisted in three patients despite programming to the lowest atrial output and/or ventricular sensitivity parameters. Cross-talk, spontaneous or provoked, has not occurred in 36 patients having currently designed 415A pulse generators manufactured with a single anodal ring in the ventricular connector channel. We conclude that: this unique form of cross-talk is "design-dependent" and occurs only in Gemini 415As with dual interconnected anodal rings and cannot occur in 415As manufactured with a single anodal ring; this form of cross-talk has occurred only in 415As with non-Cordis leads and appears to be a time-dependent development; reduction of the atrial output or use of pacing modes without atrial pacing or ventricular sensing can prevent cross-talk.(ABSTRACT TRUNCATED AT 250 WORDS)