Communication links in a wireless cellular network are subject to random slow fading. Thus a link margin is required to guarantee a certain level of link reliability. It is well known that the required link margin is reduced due to mobile handoff at the cell edge. This is the handoff gain. Traditionally the handoff gain considers only the reduction of the link margin requirement due to the presence of multiple random slow fading paths; the error rate requirement is not taken into account. However, in the case of soft handoffs, when multiple links are available to the receiver simultaneously, for a given target overall error rate, a higher error rate target per link is allowed. This increases the handoff gain, but the possibility of imbalance at the handoff links requires an additional margin. This is the differential fade margin, which reduces the handoff gain. We show how to numerically calculate the differential fade margin and the overall handoff gain without resorting to simulations. Several illustrative examples are given. The approach is used to show that the benefit of being able to allow a higher per link error rate in the soft handoff far outweighs the drawback of requiring the additional differential fade margin. We show that the soft handoff gain is more than 3 dB higher than the traditional soft handoff gain widely used in the radio frequency (RF) design community. This approach also allows us to quantify the link budget gain when the target error rate is relaxed. In addition to quantifying the gains in theoretical Gaussian and Rayleigh channels, we have also shown that in CDMA2000 3G1x systems, regardless of whether the carrier frequency band is at 1900 MHz (PCS), 850 MHz (cellular), or 450 MHz, relaxing the target frame error rate (FER) from one percent to two percent provides a 1 dB gain in the link budget, while relaxing the target FER from one percent to five percent provides a 2 dB gain in the link budget. © 2012 Alcatel-Lucent.