Operation of total heat exchangers (THXs) discussed in this article is based on thin fibrous paper, permeable to moisture but highly impermeable to air, capable of recovering both sensible and latent heat from air. The efficiencies of these exchangers are significantly influenced by air leakage from the exhaust air channel to the supply air channel, which renders traditional effectiveness definitions of heat exchangers ambiguous. The fresh air effectiveness defined on the basis of per-unit-effective-fresh-air is proposed theoretically and the leakage effect on this effectiveness is examined with experiments. Comparisons are made with the existing standards which don't include leakage effects. Experimental results also show an optimal leakage rate (~12%) in so far as the energy saving is concerned, beyond which the performance of a total heat exchanger is highly unstable. Below the threshold leakage rate, within the experiments conducted, the energy saving ratio of total heat exchanges increases with the leakage rate.