I own six books on quantum field theory (QFT): Mandl & Shaw, Peskin & Schroeder, Kaku, Maggiore, Weinberg and this one by Srednicki. Of the six, Srednicki's book was the one that finally made me feel comfortable with QFT. It's a good book for beginners, and it's also a very distinctive and unusual book. So I would definitely recommend Srednicki, and I suggest reading it alongside either Mandl & Shaw or Peskin & Schroeder.
Here are its defining features:
(1) The presentation is generally clear, elegant and encouraging. Calculations are presented in full, without glossing over intermediate steps, and the assumptions and principles behind them are made explicit (if not always entirely convincingly). There are plenty of worked examples in the text and the problems are pitched at the right level. Although I found the opening few sections a tiny bit intimidating, the reading experience was plain sailing once I got over the initial hurdle.
(2) Srednicki is also direct and to-the-point. He teaches you enough to do the calculations, and not too much more. When you are new to QFT, this is definitely what you want. But some of my friends complain that it's not thoughtful enough, and they recommend Weinberg.
(3) The chapters are each only a few pages long, and deal with just one key idea. The chapter dependencies are stated so that the student can plot his/her own way through the book. For me, this kind of organization made the material more digestible.
(4) The book is unusual in that it teaches the QFT tools through easy toy practice models before going on to algebraically demanding physical theories like QED. I found this quite effective. The downside is that you have to read 400 pages before you get to the first prediction about the real world. You can be the judge of whether you want this or not. Certainly, it's a good idea to read Srednicki alongside a book that covers things in a conventional order, like Mandl & Shaw, Peskin & Schroeder or Maggiore.
(5) Even though this is a book is an introduction, Srednicki manages to give you mini-reviews of more advanced topics: lattice QCD and confinement, chiral symmetry breaking, solitons and instantons, supersymmetry and grand unification. He doesn't say much about any of them, but what he does say helped me loads when I went on to read more specialised accounts on the individual subjects. Srednicki points you towards those specialised accounts in a bibliography at the end of every chapter. There are two topics that are conspiciously missing: QED and QCD phenomenology, and phase transitions. You need Peskin and Schroeder for that.
(6) There are two ways of doing QFT: (i) canonical quantization and time-dependent perturbation theory (the old way) and (ii) the path integral (the modern way). Srednicki adopts the path integral approach from the outset, and does it very efficiently. Nowadays, almost everyone uses the path integral approach. But the fact that Srednicki doesn't cover time-dependent pertubation theory is a problem, which is why you need to supplement your reading with either Mandl & Shaw, Maggiore or Peskin & Schroeder.
(7) Renormalization (removing infinities) is probably the hardest topic in QFT, and Srednicki's coverage of renormalization is probably the book's strongest topic. Srednicki makes renormalization feel perfectly natural; right from the beginning, he is anticipating the need for renormalization (by making a fuss about the Lehmann-Kahlen propagator and the normalization conditions in the LSZ formula). Again, I need to issue a warning. There are two equivalent ways of doing the accounting in renormalization: (i) the multiplicative approach and (ii) the counterterms approach. Nowadays, everyones uses (ii). Srednicki covers only (ii). You should get Mandl & Shaw to learn (i), and then convince yourself that (i) and (ii) are equivalent by studying Peskin & Schroeder.