Clinical Trials Meet Primary Endpoint of Non-Inferiority and Demonstrate Statistically Significant Reductions in Hypoglycemic Events and Weight Gain

Results Presented at EASD Conference in Rome

Investor Meeting and Webcast/Conference Call Scheduled for 8:00 am EDT on Tuesday, September 9, 2008

At the 44^th Annual Meeting of the European Association for the Study of Diabetes (EASD) in Rome, Biodel Inc. (Nasdaq: BIOD) today reported results from its two pivotal Phase III clinical trials designed to compare the efficacy and safety of VIAject™ to Humulin® R, a regular human insulin (RHI), in the treatment of patients with Type 1 and Type 2 diabetes. The primary objective of the trials was to determine if VIAject™ is non- inferior to RHI in the management of blood glucose levels, as measured by the mean change in patients’ glycosylated hemoglobin, or HbA1c, levels from baseline. HbA1c is a measure of average blood glucose level and an indication of how well patients are controlling their blood glucose. Based on the initial analyses, both clinical trials met the primary endpoint of non-inferior mean change in HbA1c over six months of treatment.

VIAject™ is Biodel’s proprietary injectable formulation of recombinant human insulin designed to be absorbed into the blood more rapidly than currently marketed rapid-acting insulin analogs. Two posters summarizing the preliminary Phase III clinical trials results are being presented at the EASD conference and are now available on the Biodel website, www.biodel.com. The first poster is entitled “Insulin VIAject™ and Regular Human Insulin in Patients with Type 2 Diabetes; Efficacy and Safety in an Open Label Multicenter Clinical Trial.” The second poster is entitled “Insulin VIAject™ and Regular Human Insulin in Patients with Type 1 Diabetes; Efficacy and Safety in an Open Label Multicenter Clinical Trial.”

Type 2 Clinical Trial

The Type 2 trial was an open-label, parallel group, randomized study conducted at 47 centers in the United States, Germany, and India. In this trial, 186 patients receiving VIAject™ and 205 patients receiving RHI were evaluable, and HbA1c decreased comparably in the two treatment groups with an adjusted difference between the groups of -0.2%. With a 95% confidence interval of -0.4 to -0.1, non-inferiority between VIAject™ and RHI was established.

Type 2 patients receiving VIAject™ experienced statistically significant (p<0.01) fewer mild and moderate, or non-severe, hypoglycemic events than patients receiving RHI. The percentage of patients experiencing one or more mild and moderate hypoglycemic events was 67.2% for those receiving VIAject™ and 71.7% for those receiving RHI, representing a 6.0% lower rate. The rate of mild and moderate hypoglycemic events per treatment month was 1.5 for patients receiving VIAject™ and 2.3 for patients receiving RHI, a difference of 35% favoring patients who received VIAject™. The total number of mild and moderate events was 1,566 for VIAject™ patients and 2,783 for RHI patients.

Type 2 patients receiving VIAject™ gained an average of 0.3 kg (0.7 lbs) over 26 weeks compared to patients receiving RHI, who gained an average of 1.7 kg (3.7 lbs), for a net adjusted difference of 1.4 kg (3.1 lbs). This difference was statistically significant (p<0.01).

Mean insulin antibody level changes were comparable between treatment groups. The mean change from baseline was 0.2 +/- 3.6 U/ml in the VIAject™ treatment group and 0.3 +/- 1.8 U/ml in the RHI treatment group.

Type 1 Clinical Trial

The Type 1 trial was an open-label, parallel group, randomized study conducted in 60 centers in the United States, Germany, and India. The primary efficacy analyses (HbA1c, hypoglycemia, weight) were based upon 131 patients who received VIAject™ and 140 patients who received RHI in the United States and Germany. In these analyses, HbA1c decreased comparably in the two treatment groups with an adjusted difference between the groups of -0.1%. With a 95% confidence interval of -0.3 to +0.1, non-inferiority between VIAject™ and RHI was established.

A statistically significant (p<0.01) interaction associated with data from India was observed and efficacy results from India are, therefore, not comparable to the results from the United States and Germany. Due to this interaction, the efficacy data from India are not included in the preliminary analyses. These data are not comparable to the rest of the data for several reasons including: (a) markedly increased HbA1c levels both prior to and after study drug initiation, (b) medically improbable doubling of intra-subject variability in HbA1c results both prior to and after study drug initiation and (c) markedly reduced reporting of any hypoglycemic events.