PURPOSE The ability to detect small malignant lesions with magnetic resonance imaging (MRI) is limited by inadequate accumulations of Gd with standard chelate agents. To date, no T1-targeted agents have proven superiority to Gd chelates in their ability to detect small tumors at clinically relevant field strengths. Activatable cell-penetrating peptides and their Gd-loaded dendrimeric form (ACPPD-Gd) have been shown to selectively accumulate in tumors. In this study we compared the performance of ACPPD-Gd vs. untargeted Gd chelates to detect small tumors in rodent models using a clinical 3T-MR system. MATERIALS AND METHODS This study was approved by the Institutional-Animal Care-and-Use Committee. 2 of 4 inguinal breast fat pads of 16 albino-C57BL/6 mice were inoculated with tumor Py8119 cells and the other 2 with saline at random. MRI at 3T was performed at 4, 9, and 14 days after inoculation on 8 mice 24-hours after injection of 0.036mmol Gd/kg (ACPPD-Gd), and before and 2-3 minutes after 0.1 mmol/kg gadobutrol on the other 8 mice. T1-weighted (T1w) tumor signal normalized to muscle, was compared among the non-contrast, gadobutrol, and ACPPD-Gd groups using ANOVA. Experienced and trainee readers blinded to experimental conditions assessed for the presence of tumor in each of the 4 breast regions. Receiver operator characteristic (ROC) curves and area-under-curve (AUC) values were constructed and analyzed. RESULTS Tumors ≥1mm3 were iso-intense to muscle without contrast on T1w sequences. They enhanced diffusely and homogeneously by 57±20% (p<0.001) 24 hours after ACPPD-Gd and by 25±13% (p<0.001) immediately after gadobutrol. The nearly 2-fold difference was similar for small tumors (1-5mm3) (45±19% vs. 19±18%, p = 0.03). ACPPD-Gd tended to improve tumor detection by an experienced reader (AUC 0.98 vs 0.91) and significantly more for a trainee (0.93 vs. 0.82, p = 0.02) compared to gadobutrol. This improvement was more pronounced when obvious tumors (>5mm3) were removed from the ROC analysis for both the experienced observer (0.96 vs. 0.86) and more so for the trainee (0.86 vs. 0.69, p = 0.04). CONCLUSION ACPPD-Gd enhances MMP-expressing tumors of any size at 3T 24 hours after administration, improving their detection by blinded observers when compared to non-contrast and contrast groups given commercial Gd-chelates and imaged during the equilibrium phase.