Targeting the gastrin-releasing peptide receptor (GRPR) with the bombesin analogue RM26, a 9 aa peptide, has been a promising strategy for cancer theranostics, with recent success in radionuclide imaging of prostate cancer. However, therapeutic application of the short peptide RM26 would require a longer half-life to prevent fast clearance from the circulation. Conjugation to an albumin-binding domain (ABD) is a viable strategy to extend the in vivo half-life of peptides and proteins. We previously reported an ABD-fused RM26 peptide targeting GRPR (ABD-RM26 Gen 1) that showed prolonged and stable tumor uptake over 144 h; however, the observed high kidney uptake indicated that the conjugate’s binding to albumin was reduced and that this could be an obstacle for its use as a delivery system for targeted therapy, especially for radiotherapy. Here, we have designed, produced, and preclinically evaluated a series of novel ABD-RM26 conjugates with the aim of improving the conjugate’s binding to albumin and decreasing the kidney uptake. We developed three second-generation constructs with varying formats, differing in the relative positions of the targeting moieties and the radionuclide chelator. The produced conjugates were radiolabeled with indium-111 and evaluated in vitro and in vivo. All constructs displayed improved biophysical characteristics, biodistribution, and lower kidney uptake compared to previously reported first-generation molecules. The ABD-RM26 Gen 2A conjugate showed the best biodistribution profile with a nearly 6-fold reduction in kidney uptake. However, the ABD-RM26 Gen 2A conjugate’s binding to GRPR was compromised. This conjugate’s assembly of albumin- and GRPR-binding moieties might be used for further development of drug conjugates for targeted therapy/radiotherapy of GRPR-expressing cancers.