The influence of varied linear array configurations on passive ranging is determined when imperfect measurements of the time delays between the sub‐arrays are available. The exact bias and farfield bias and variance relations on the source range estimates are derived as a function on unequal sub‐array separation. The results are contrasted with those found in the literature on equally separated sub‐arrays to highlight the sensitivity and resulting degradation in performance. It is found, for instance, that the range bias more than doubles when the internal sub‐array is moved from the midpoint to the quarter‐point from either end sub‐arrays and more than quadruples when it is moved from the quarter‐point to the one‐eighth‐point. Since the range variance is related to the true range and to the incurred bias, it degrades much more rapidly. The commonality between array ranging and multipath ranging is also developed for various channels of interests. The optimal placement for the receiving sensor in multipath channels is derived. Though it turns out that the same relations apply to both ranging problems, the existence of a multipath channel will support a longer ranging capability with the baseline now effectively determined by such large parameters as water depth, receiver, and source depths. Such results enable the merger of multisensor data according to their individual merits and the provision for a credence estimate of the source range choice.